Learning and Individual Differences 29 (2014) 98–105

Contents lists available at ScienceDirect

Learning and Individual Differences

journal homepage: www.elsevier.com/locate/lindif

Systemizing and special interests: Characterizing the continuum

from neurotypical to autism spectrum disorder
Catherine L. Caldwell-Harris ⁎, Chloe J. Jordan
Department of Psychology, Boston University, 64 Cummington St., Boston, MA 02215, USA

a r t i c l e i n f o a b s t r a c t

Article history: Special interests have been studied in children with autism spectrum disorder (ASD) but not in adults. Using an
Received 8 September 2012 online survey, it was found that individuals with ASD reported more intense interests in systemizable domains,
Received in revised form 25 July 2013 relative to neurotypical adults. Self-reported systemizing preference was correlated with intensity of interest in
Accepted 11 October 2013
systemizable domains both for those with ASD and for neurotypical young adults. Few gender differences were
found in the neurotypical group in the expected categories of machines, technology and vehicles, where gender
Keywords:
Autism spectrum disorder
differences have been found in children. Gender differences in these categories did appear for the ASD group. We
Asperger syndrome propose a strength-based model of special interests, with the hobbies of neurotypical forming a continuum with
Special interests the special interests of ASD.
Adults © 2013 Elsevier Inc. All rights reserved.
Systemizing

1. Introduction
Among individuals with Asperger syndrome, repetitive behaviors
often occur in the form of intense interests which may be considered
unusual or atypical in their content (e.g., World War II airplanes; see
Asperger, 1991; Attwood, 2003; Winter-Messiers, 2007). These are
sometimes called restricted or circumscribed interests (e.g., Klin,
Danovitch, Merz, & Volkmar, 2007). Following Winter-Messiers (2007),
we use the term “special interests” because it promotes a strength-
based approach and it is also the term most frequently used by
individuals on the Internet discussion forums from which we recruited
survey participants in the current study. We describe and defend
the view that special interests are not primarily repetitive behaviors,
but reflect information processing styles and cognitive strengths
(Baron-Cohen, 2002). The interests and hobbies of neurotypical
individuals (i.e., those without ASD, Attwood, 1998; Winter-Messiers,
2007) frequently reflect individuals' cognitive-personality styles, and
we thus propose that the special interests of those with Asperger
syndrome or on the autism spectrum lie on a continuum with
neurotypical hobbies. This leads to two primary hypotheses: (1) that
the content of special interests exists on a continuum between
neurotypical individuals and individuals with autism spectrum disorder
(ASD), and (2) that special interests reflect information processing styles,
such that the interests of individuals with ASD are correlated with
systemizing ability, and interests of neurotypicals with mentalizing
ability, as measured by the Systemizing Quotient (Wheelwright et al.,
⁎ Corresponding author.
E-mail address: charris@bu.edu (C.L. Caldwell-Harris).
2006) and the Reading the Mind in the Eyes Test (Baron-Cohen,
Wheelwright, Hill, Raste, & Plumb, 2001).
Past research has focused primarily on the special interests of
children with autism spectrum disorder (ASD), who develop these
interests as early as 1–4 years of age (Attwood, 2003; Bashe & Kirby,
2001; Moore & Goodson, 2003), although some interests cannot be
reliably scored on certain assessments until age 4 (Moore & Goodson,
2003). Approximately 75–90% of individuals with mild to moderate
autism or Asperger syndrome develop one or more special interests
(Bashe & Kirby, 2001; Klin et al., 2007). Other individuals on the autism
spectrum, including those with PDD-NOS (Sturm, Fernell, & Gillberg,
2004) and Rett syndrome (Mazzocco et al., 1998), also exhibit intense
interests.
Special interests are often manifested in efforts to collect objects or
information relevant to the interest area, which can require extensive
amounts of time (Bashe & Kirby, 2001). Parents often find special
interests the most difficult to accommodate of autistic behaviors due
to their intensity (Mercier, Mottron, & Belleville, 2000). Unlike other
behaviors, special interests do not lessen with age (Fecteau, Mottron,
Berthiaume, & Burack, 2003) and the number of interests may increase
as the individual approaches adulthood (Bashe & Kirby, 2001).
Among children with ASD, special interests often reflect exceptional
abilities, such as systemizing and heightened attention to detail.
Systemizing is the drive to explore, analyze, and construct systems
(Baron-Cohen, Richler, Bisarya, Gurunathan, & Wheelwright, 2003).
Domains that are amenable to systemizing are rule-based and
predictable, facilitating detection of input-operation–output relationships
(Baron-Cohen, Ashwin, Ashwin, Tavassoli, & Chakrabarti, 2009). The
special interests of children with ASD frequently occur in systemizable

1041-6080/$ – see front matter © 2013 Elsevier Inc. All rights reserved.
http://dx.doi.org/10.1016/j.lindif.2013.10.005
 C.L. Caldwell-Harris, C.J. Jordan / Learning and Individual Differences 29 (2014) 98–105
domains, such as in mechanical, collectible, natural, and numerical
systems (Baron-Cohen & Wheelwright, 1999).
Neurotypical (NT) individuals also develop special interests. While
up to 90% of individuals with ASD develop special interests (Bashe &
Kirby, 2001; Klin et al., 2007), about 30% of NT children develop an
“extremely intense interest” (DeLoache, Simcock, & Macari, 2007).
Special interests in NT children may appear as early as 1–2 years of
age. Like children with ASD, the special interests of NT children reflect
their cognitive strengths. NT children are often preoccupied by interests
in people, imaginative play, and the social environment more broadly
(Attwood, 2003; DeLoache et al., 2007). These interests are consistent
with strong mentalizing ability (i.e., the capacity to understand and
attribute mental states to others, Baron-Cohen, Wheelwright, Hill,
Raste, & Plumb, 2001; Frith, Morton, & Leslie, 1991). While individuals
with ASD often have higher systemizing ability than NT individuals
(Baron-Cohen et al., 2003), NT individuals generally have higher
mentalizing abilities (Baron-Cohen, Wheelwright, Hill, et al., 2001;
Baron-Cohen, Wheelwright, Skinner, Martin, & Clubley, 2001).
Furthermore, NT females tend to have higher mentalizing ability than
NT males (Baron-Cohen, Wheelwright, Hill, et al., 2001; Baron-Cohen,
Wheelwright, Skinner, et al., 2001), while NT males generally have
higher systemizing ability than NT females (Baron-Cohen et al., 2003).
Little is currently known about special interests in adulthood.
Understanding adults' special interests is important for illuminating the
positive role that interests play in adult life, such as rewarding careers
(Attwood, 2003; Grandin & Duffy, 2008; Jackson, 2002). No studies yet
compare the interests of individuals with ASD to those of NT individuals.
The current study examines the content of special interests held by
adults with ASD and NT adults, with the expectation that individuals
with ASD will hold more intense interests in domains previously
identified as systemizable, such as machines and technology, sorting,
categorizing, and organizing, factual and numerical information,
collecting, and the sciences. We also expect that NT individuals will
hold more intense interests in mentalizing domains, such as in people
and sports and games (as discussion of sports fosters social affiliation,
see Mueller, Agamanolis, & Picard, 2003). The current study will also
assess participants' systemizing and mentalizing abilities, with the goal
of identifying whether interest categories previously identified as
systemizable correlate with high systemizing ability, and likewise
whether interest categories expected to rely on mentalizing correlate
with high mentalizing ability. By identifying cognitive abilities that
contribute to special interests, we hope to develop a strength-based
continuum model of special interests, where the interests of individuals
with ASD and NT individuals coexist and vary depending on systemizing
and mentalizing ability (see Fig. 3 and Discussion section). A strength-
based model of special interests can inform the development of
educational and therapeutic programs that capitalize on individual
abilities and use special interests as a medium for learning other
important skills.
We also hope to better understand gender differences in special
interests. Winter-Messiers (2007) reported that ASD girls were more
likely than boys to have more neurotypical interests, such as a
10 year-old girl being interested in horses. Attwood (2003) also noted
ASD girls' intense interests in dolls, animals and fiction, but remarked
on some aberrant qualities of these interests. Girls didn't use their
dolls to play with others, but would self-play with a large doll collection.
An interest in animals could become consuming; girls would want to
act like animals, or to sleep in a stable. Interests in fiction included
collecting and wanting to contact the author. The focus of the writing
interest was not related to school success or an anticipated career.
ASD is well known to be less frequent in girls, which has
been hypothesized to reflect protective aspects of girls' socialization
(e.g., Constantino & Todd, 2003) or lower fetal testosterone (Auyeung
et al., 2006). It would be useful to know if systemizing and/or a
person's autism traits predicted their type of interests independently
of gender.
99
Gender differences in interests also exist in typically developing
children. In the study of DeLoache et al. (2007), half of the intense
interests that parents reported for boys were for vehicles, trains, and
machines, and another 27% were for balls, dinosaurs, and tools. Girls'
intense interests were clothes/dressing up, babies, and tea sets. These
gender differences in interests were observed in the youngest children
studied (approximately 1 year old). Since knowledge of gender stereo-
types, implicit or explicit, does not occur until 18 months or later,
DeLoache et al. speculated that biological factors, such as fetal
testosterone, play a role.
Gender differences in special interests have not previously been
studied in adult persons with ASD. Our secondary hypothesis is that
we will see continuity between childhood and adulthood. We thus
expect that both ASD and neurotypical males will have interests in
more systemizable domains than will ASD and neurotypical females.
Understanding gender differences in special interests during adulthood
is important, as it can provide a means to adapt skill development in the
workplace and higher educational settings to individual cognitive styles.
2. Materials and methods
2.1. Participants
Participants with ASD were recruited from WrongPlanet.net, a large
online discussion forum created for individuals with ASD, and their
families, friends, and supporters (Plank & Yellow Sneaker Media,
2004). Among 70 individuals who self-reported themselves as having
an ASD, 67 identified themselves as having Asperger syndrome, while
only 2 reported autism spectrum disorder and 1 reported PDD-NOS.
Slightly more than half (58.6%) of individuals with ASD reported being
diagnosed by a clinician (see Table 1 for summary of other demographic
and diagnostic variables). The most common disorders reported by both
groups were anxiety disorders (including panic attacks, social phobia,
generalized anxiety), followed by depression. Less commonly reported
was ADHD, followed by OCD.
In the ASD group, 42.9% of participants were female, while 50.4% of
the NT group was female. Although ASD is typically more prevalent in
males, this gender ratio is similar to that reported in past studies of
ASD special interests on Internet discussion forums (Jordan & Caldwell-
Harris, 2012).
Neurotypical participants included participants who were forum
users from the WrongPlanet.net website (n = 68) who identified
themselves as neurotypical (i.e., not having an ASD) and students
from Boston University who participated for course credits (n = 51).
Recruitment from a university allowed us to approximately match the
age and education level of the ASD and NT groups.
To determine whether participants reporting themselves as NT who
were recruited from the WrongPlanet.net forum should be considered
neurotypical, we compared their Systemizing Quotient, Autism
Quotient, and Eyes Test scores (see Materials) to those earned by NT
participants recruited from Boston University. WrongPlanet NT
individuals did score slightly higher on the autism quotient than Boston
University NT participants (WrongPlanet NT M = 22.9, SD = 11.1, BU
M = 17.1, SD = 7; t(79) = −3.1, p b .005) and thus some of them may
be considered part of the broader autism phenotype (Constantino &
Todd, 2003). However, there were no differences between the two NT
groups on the Systemizing Quotient or Eyes Test. Furthermore,
WrongPlanet NT individuals had lower scores than ASD individuals on
the Autism Quotient (WrongPlanet NT M = 22.9, SD = 11.1, ASD M =
36.4, SD = 7.3; t(78) = −7.2, p b .001) and the Systemizing Quotient
(WrongPlanet NT M = 35.1, SD = 14.7, ASD M = 43.8, SD = 16.5;
t(97) = −2.8, p b .05), and higher scores on the Eyes Test (WrongPlanet
NT M = 24, SD = 5.4, ASD M = 18.4, SD = 9; t(88) = 3.9, p b .001).
Although the WrongPlanet NT participants were substantially
neurotypical, their scores were slightly elevated compared to the Boston
University NT students and the NT norms of 16.4 reported by Baron-
100 C.L. Caldwell-Harris, C.J. Jordan / Learning and Individual Differences 29 (2014) 98–105

Table 1
Summary of participants.

Age M Male % Ethnicity Education % Clinician % Receiving % With other If have other disorder,
(SD, range) (% non-Caucasian) (% some college) diagnosed ASDa therapy for ASD (non-ASD) disorder % clinician diagnosedb

Autism spectrum disorder 29.2 (12.5, 16–42) 57.1% 4.2% 74.2% 58.6% 67.1% 68.6% 87.5%
(N = 70)
Neuro-typical 23.8 (10.7, 15–59) 49.6% 23.4% 80.6% – – 22.7% 88.9%
(N = 119)

Note. ASD = autism spectrum disorder.

a
The remaining percentage were self-diagnosed (11.4%) or did not respond (30%).
b
The remaining percentage were self-diagnosed (6.3% of ASD group, 0% of NT group) or did not respond (6.3% of ASD group, 1.1% of NT group).

Cohen, Wheelwright, Hill, et al. (2001), Baron-Cohen, Wheelwright, 3. Results

Skinner, et al. (2001). Our overall NT sample thus did have a slightly
higher than average number of autistic traits. 3.1. Preliminary analyses Systemizing Quotient and Eyes Test

2.2. Materials 3.1.1. Internal consistency of scales

An online survey was administered using SurveyMonkey online
survey software. The Cambridge University Obsessions Questionnaire
(Baron-Cohen & Wheelwright, 1999) was modified by adding the
categories of nature, history and culture, and psychological disorders,
with the existing category of plants being included under the broader
nature category (see Table 2).
If a participant indicated an interest in a given category, they were
asked to provide a brief description of their particular interest, and to
rate the intensity of their interest in from 1 to 3, 1 indicating a casual
interest, 2 a moderate interest, and 3 an intense interest. Participants
who did not express an interest in a given category were assigned an
interest intensity of 0.
To measure how strongly autistic traits were expressed, the
survey included the Autism Spectrum Quotient (AQ; Baron-Cohen,
Wheelwright, Skinner, et al., 2001). The Reading the Mind in the Eyes
Test Revised (Eyes Test, Baron-Cohen, Wheelwright, Hill, et al., 2001)
provided a measure of mentalizing ability, and a modified version of
the Systemizing Quotient-Revised (SQ, Wheelwright et al., 2006) as-
sessed systemizing ability.
The internal consistency of the Autism Spectrum Quotient (AQ) was
high, with a Cronbach alpha of 0.94. Our modified version of the
Systemizing Quotient-Revised (SQ-R) also showed high consistency,
with a Cronbach alpha coefficient of 0.89.
3.1.2. Effects of age and diagnosis type
The Systemizing Quotient, AQ, and Eyes Test were examined for
effects of age and whether individuals reported a disorder other than
ASD. Age did not correlate significantly with scores on any of the three
measures, nor were there differences in scores between individuals
with a disorder other than ASD in either group. In the ASD group, no
differences in scores on these three scales were obtained for individuals
diagnosed by a clinician and those who were self-diagnosed, or between
individuals who had received therapy and those who had not.
3.1.3. Correlations between scales
Consistent with prior work, the Systemizing Quotient correlated
positively with the AQ (r = 0.49, p b .001), and negatively with the
Eyes Test (r = −.19, p b .05). The AQ also correlated negatively with
the Eyes Test (r = −.49, p b .001).

Table 2
Special interest categories.

Special interest category Description

Machines & technology Interests in computers, radios, television, clocks, etc., or expresses interest in how things work. Does not include interests in
watching or listening to movies, television, music, etc.
Belief systems Interests in religions or mythologies, political systems, philosophies, alternative beliefs (such as conspiracy theories), etc.
Numerical information Interests in timetables, calculators or calculations, calendars, prime numbers, reading or creating charts or tables of information, etc.
Sports & games Interests in football, tennis, walking, biking, tennis, playing cards, chess, board games, puzzles, video games, etc.
Item attachment Interests focusing on a particular item or type of object, as well as interests in certain words or phrases.
Sensory Interests in touching or feeling certain things, or mentions fascinations with texture, specific sounds, lighting, colors, smells, etc.
Crafts Interests in model making, knitting, sewing, carpentry, etc.
Factual information Interests in reading or memorizing lists, reading encyclopedias, dictionaries, newspapers, etc.
Creative arts Interests in movies, television shows, art work, painting, playing an instrument, music, writing and reading fiction, creating media
(e.g., online films), performing arts, etc.
Psychological disorders Expresses interest in any psychological disorder.
History & culture Interests in existing languages, particular countries or civilizations, time periods or eras in history, etc.
Information & mechanical systems Interests in plumbing, light switches or electrical wiring, maps, city planning, subway maps and/or schedules, businesses and organizations,
etc., or an interest in creating systems (such as languages or maps).
Sorting, categorizing, & organizing Interests in making lists, lining objects up, arranging objects in certain orders or categories, planning, or obsession with neatness or organization.
Food & drink Interests in consuming or creating a particular food or drink, cooking, baking, etc.
Vehicles Interests in trains, airplanes, buses, boats, cars, etc.
People Interests in a particular person, in types or groups of people, or in interacting with people, including participation in online communities.
Collecting Interests in acquiring collections of particular items, e.g., bottles, keys, caps, stamps, rocks, etc.
Nature Interests in plants, interacting with nature (e.g., hiking, gardening, exploring, etc.), and natural phenomena
(e.g., volcanoes, tsunamis, lightning, etc.).
Animals Interests in pets, wild or farm animals, insects, fish, birds, etc., also includes animal-related activities (e.g., bird watching), but does not include
mythical creatures.
Sciences Interests in astronomy, chemistry, biology, physics, engineering, mathematics, logic, economics, psychology, meteorology,
specific diseases or conditions, etc.
Other Interests that did not fit within a specific category.
 C.L. Caldwell-Harris, C.J. Jordan / Learning and Individual Differences 29 (2014) 98–105 101

3.1.4. Gender and group differences in the scales Table 4

For each of the three measures, a two-way between-groups analysis Intercorrelations for the intensity of interest in special interest categories and scores on the
Systemizing Quotient and Eyes Test.
of variance was conducted to examine main effects of gender and
group as well as potential interaction effects. Not surprisingly, ASD Special interest category Systemizing Quotient Eyes Test
participants had significantly higher scores than NT participants on the ASD Neurotypical Combined ASD
Systemizing Quotient (ASD M = 44.8, SD = 16.8, NT M = 32.4, SD = only only AD and NT only
11.9, F(1, 108) = 14.1, p b .001, η2p = .12) and on the AQ (ASD M = 36.4, Machines & technology .30# .42⁎⁎ .40⁎⁎⁎ −.08
SD = 7.3, NT M = 20.0, SD = 9.7, t(153) = −11.2, p b .001, η2 = .45). Vehicles .47# −.21# .30⁎⁎ 0
Also as hypothesized, NT participants had higher Eyes Test scores than Sorting, categorizing, & organizing .13 .11 .17# −.02
ASD participants (ASD M = 18.4, SD = 9, NT M = 24.9, SD = 5, Item attachment .02 .27# .25⁎⁎ −.10
Sensory .08 .17 .19# −.11
t(71.8) = 4.8, p b .001, η2 = .14). The only gender differences occurred Factual information .36⁎⁎ .35⁎⁎ .41⁎⁎⁎ −.05
for the Systemizing Quotient, with no interaction of gender and Collecting .29# .17 .32⁎⁎⁎ −.01
diagnosis. Males had higher systemizing scores than females (male Sciences .27# .05 .18# .20
M = 42.6, SD = 15.5, female M = 33.6, SD = 13.8, F(1, 108) = 4.8, Food & drink .08 .14 −.01 .08
Average interest intensity .38⁎⁎ .38⁎⁎ .37⁎⁎ .23#
p b .05, η2p = .04).
Note. For the Systemizing Quotient, higher scores are indicative of higher systemizing
3.2. Special interests analyses ability, while for the Eyes Test, higher scores are indicative of higher mentalizing ability.
Participants indicated the intensity of their interests on a scale from 0 to 3, 0
corresponding to no interest, 1 a casual interest, 2 a moderate interest, and 3 an intense
Our primary hypotheses were that individuals with ASD would have interest.
interests of different contents than NT individuals, and that interests ⁎⁎⁎ p b .001.
would vary with systemizing and mentalizing abilities for both groups. ⁎⁎ p b .005.
#
p b .05.
As shown in Table 3, individuals with ASD had more intense interests
in item attachment, collecting, factual information, and sorting,
categorizing, and organizing, as well as in sciences, sensory, and 3.3. Special interests, group and gender
machines and technology. NT individuals had more intense interests in
food and drinks. Individuals with ASD also identified interests in more Our secondary hypothesis was that ASD and NT males would have
categories overall than NT individuals (ASD M = 10.9, SD = 4.9, NT interests in more systemizable domains than ASD and neurotypical
M = 9, SD = 5.2, t(187) = −2.4, p b .05, η2 = .03), and had more intense females. We analyzed how gender interests interact with diagnosis and
interests on average (ASD M = 1.3, SD = 0.7, NT M = 1, SD = 0.6, type of special interest by conducting 2 × 2 analyses of variance on each
t(187) = −2.8, p b .01, η2 = .04). interest category. Four categories revealed significant 2 × 2 interactions.
Combining over both ASD and neurotypical individuals, scores on The resulting patterns are complex and can be best appreciated in
the Systemizing Quotient correlated positively with intensity of interest graphic forms. Fig. 1 plots the significant gender X diagnostic group
in machines and technology, sorting, categorizing, and organizing, item interactions for machines and technology, η2 = .04, and vehicles, η2 =
attachment, sensory, and factual information, as well as vehicles, .05, both ps b .05. These are plotted together because they show the
collecting items, and sciences (see Table 4). Scores on the Eyes Test same pattern of no gender differences for neurotypicals, but males having
correlated positively with food and drink, but negatively with item stronger interests than females in the ASD group, confirmed with post-
attachment, factual information, and collecting. The Systemizing hoc Scheffe tests. Fig. 2 plots the significant interactions for sorting,
Quotient also correlated positively with the total number of interest categorizing, and organizing, η2 = .06, and collecting interest category,
categories (M = 9.7, SD = 5.1, r = .37, p b .001) and with the average η2 =.05, both psb.05. We plotted these special interests together because
intensity of all interests (M = 1.1, SD = 0.7, r = .41, p b .001). Many the Scheffe post-hoc tests revealed a pattern of greater interests by
correlations remained significant even when correlating within the females than males among neurotypicals for both interest categories,
ASD or neurotypical group, although correlations were lower because but either no gender differences in the ASD group (Sorting, Organizing)
the SQ range was smaller when only one group was analyzed. Finding or males having somewhat greater interest than females (but smaller
robust correlations between systemizing scores and intensity of differences than seen in the ASD group in Fig. 1).
interests among neurotypicals is consistent with the concept of a Since the lack of gender differences among neurotypicals was
continuum between neurotypicals' hobbies and ASD individuals' special unexpected, we scrutinized individuals' comments accompanying
interests. their intensity ratings to determine whether engagement with these
objects occurred because objects facilitate systemizing.
Table 3
Intensity of special interests of individuals with autism spectrum disorder and 3.3.1. Machines and technology
neurotypical individuals.
Some comments did indicate that the object afforded systemizing, as
Special interest category Autism spectrum Neurotypical t(187) η2 in examples (1)–(3):
disorder M (SD)
M (SD)
(1) “I like to take things apart when I get the opportunity to observe
Item attachment 1.5 (1.3) 0.7 (1.2) −4.1⁎⁎⁎ .08 how it functions firsthand.” Female NT
Collecting 1.1 (1.2) 0.6 (0.9) −3.3⁎⁎⁎ .06 (2) “Building computers from scratch and this hobby/interest has
Factual information 1.8 (1.3) 1.1 (1.2) −3.5⁎⁎⁎ .06
been around since I was 8 years old and its still around.” Male
Sorting, categorizing, & organizing 1.7 (1.1) 1.2 (1.2) −2.7⁎⁎ .04
Sciences 1.6 (1.3) 1.3 (1.3) −2.6⁎⁎ .04 ASD
Sensory fixation 1.7 (1.1) 1.3 (1.3) −2.4# .04 (3) “Computers, mostly. Have repaired and built several. Also a Linux
Machines & technology 1.6 (1.4) 1.2 (1.2) −2.3# .03 user.” Female NT
Food & drink 0.3 (0.5) 0.5 (0.5) 3.6# .06
However, comments such as those in (4)–(6) emphasized machines
Note. Participants indicated the intensity of their interests on a scale from 0 to 3, 0
corresponding to no interest, 1 a casual interest, 2 a moderate interest, and 3 an intense as merely conduits to information, not as being mechanistically
interest. interesting.
⁎⁎⁎ p b .001.
⁎⁎ p b .01. (4) “Tv since I was 4, I can watch tv for hours at a time. I am now
#
p b .05. addicted to the Internet.” Female NT
102 C.L. Caldwell-Harris, C.J. Jordan / Learning and Individual Differences 29 (2014) 98–105

Fig. 1. Respondents' intensity of interests (0–3 scale) is plotted for two categories that both show a similar interaction between gender and group (ASD vs. neurotypical). Bars show that
ASD males have stronger interests than females in the interest category of machines and technology, and also the interest category of vehicles. Neurotypical males and females did not
differ.

(5) “I am in college so obviously my laptop, ipod, and other
electronics are basics in my life!” Female NT
(6) “I like using my Iphone and Mac for virtually everything.” Male NT
The proportion of males and females who gave examples of each
type were similar, although the comments that emphasized systemizing
were fewer in number in both genders.
3.3.2. Vehicles
ASD males reported the highest intensity of interests in vehicles, but
the average intensity was still only just above 1, indicating a casual
interest, while ASD females were almost 0 (on average no interest).
While NT males and females reported having similar intensity of
interests, both averaged around 0.5, which is low compared to other
interest categories. Lack of gender differences may thus have occurred
because of floor effects. But a larger issue has become apparent,
concerning these interests and developmental stage. Vehicles (like
machines) afford systemizing because they have many parts that can
be analyzed. These parts can be studied to determine how they work
as independent components (doors and windows open and close,
wheels turn), and what these parts contribute to the functioning of
the whole object. Vehicles also afford systemizing because sets of
vehicles can be compared and contrasted, thus yielding opportunities
for developing mechanistic theories around attributes such as the
function and speed of different categories of vehicles. But note that
these are reasons why children who like systemizing will be attracted

Fig. 2. Respondents' intensity of interests (0–3 scale) is plotted for two categories, in order to depict the interaction between gender and diagnostic group (ASD vs. neurotypical). Bars show
that ASD males have stronger interests than females in the interest category of collecting, but were only slightly (and nonsignificantly) more interested in sorting/collecting. For both of
these categories, among neurotypical respondents, females expressed more intense interests than males.
 C.L. Caldwell-Harris, C.J. Jordan / Learning and Individual Differences 29 (2014) 98–105
to vehicles. Adults who like systemizing will turn to domains that are
more challenging for their cognitive abilities (computer programming,
the sciences).
3.3.3. Sorting, categorizing and organizing
The comments of female neurotypicals were similar to ASD males
and females. These three groups also had roughly similar reported
intensities (see Fig. 2). Examples appear in (7)–(9). By contrast, NT
males expressed almost no interests in this category.
(7) “I like to write things down in my yearly organizer, if I have any
plans. Only when cleaning my room, I sometimes like to arrange
objects in a certain order.” ASD female
(8) “Since I was a child, I have loved organizing and planning.
Sometimes I spend a whole day to clean my room. Especially,
my closet has to be in some order. Ex. colors.” NT female
(9) “I *love* list-making; lining objects up; symmetry; organizing
music; filing; data entry.” ASD male
The comments overall suggest that sorting, categorizing and
organizing inhabit the shallow level of systemizing. They are not used
to analyze or predict a system, but to impose control and order into
one's life. There was a trend among the comments for all females to
sort, categorize and organize as an end towards a goal of having an
organized life, while ASD males reported enjoying sorting, categorizing
and organizing because it enabled systemizing (i.e., understanding)
systems that were not set up by oneself.
3.3.4. Collecting
Neurotypical respondents reported quite low intensities, with only
ASD males reporting on average having a moderate interest in
collecting. Types of collections were similar for neurotypical males and
females and ASD females, with rocks being the most common item
collected, then shot glasses, sea shells, and coins, as in examples
(10)–(12).
(10) “Rocks and minerals, turtles (not real ones), stamps, coins,
books, …spool knitters, small looms, antique plates…” NT Female
(11) “Dolls, buttons, fairies, egyptian, chinese, flying tigers” ASD
Female
(12) “Collect boxes and perfume oils.” ASD Female
Statements were more eccentric for ASD males, as in (13)–(14).
(13) “When I was 4, I used to collect the cushions that go on the gaps
between eye pieces of swimming goggles. At various times in
childhood I obsessed over certain items, but this no longer
happens.” ASD Male
(14) “I delete nothing. Not files, not letters, not notes, not music, not
books, not computers, not electronics, nothing. I keep it all…”
ASD Male
This category thus resembles the vehicles interest in that floor
effects in interest intensity means reduced opportunity to observe
gender effects. The higher interest levels reported among ASD males
are consistent with this group having the highest systemizing
propensity but also may reflect hoarding, a characteristic of OCD,
which can be co-morbid with ASD, as in example (14).
4. Discussion
4.1. Gender differences and special interests
We found no gender differences in interests among neurotypical
adults, contrary to the robust differences observed in children
(DeLoache et al., 2007). While this could reflect lack of power and
sampling error due to small sample sizes, we suggest that this is at
103
least partly an artifact of using The Cambridge University Obsessions
Questionnaire (Baron-Cohen & Wheelwright, 1999; see Table 2). Before
widespread computer and Internet use, being interested in machines
and technology typically occurred because machines and technology
afforded systemizing (taking apart machines; computer programming).
But the neurotypical young adults we surveyed reported being
interested in machines and technology primarily for information access,
socializing, and entertainment. No gender differences were suggested
by the few respondents who did report having systemizing interests
in machines and technology. Similarly, lack of gender differences in
interest in vehicles, and possibly also in sorting, organizing, categorizing
and collecting may have occurred because these interests do not afford
challenging systemizing opportunities for adults. Future work on
gender differences in systemizing can start anew by asking what
systemizing activities are common for the target age group in the
Internet and personal computer era.
4.2. What causes special interests?
The prevailing explanations for special interests are (1) that they
function to reduce anxiety (Attwood, 2003), or (2) reduce over-
arousal caused by sensory stimuli (Hutt, Hutt, Lee, & Ounsted, 1965).
These explanations need to be reconsidered given Turner's (1997)
observation that individuals with ASD engaged in repetitive behaviors
and special interests when alone and un-stimulated. Turner also found
that repetitive behaviors were least likely to occur during social
interactions when anxiety is maximal.
We propose that people's unique cognitive abilities, and especially
systemizing and mentalizing, prompt the development of special
interests in particular domains (as illustrated in Fig. 3; see also Baron-
Cohen & Wheelwright, 1999). In this view, ASD special interests exist
on a continuum with NT interests. ASD special interests that draw on
strong systemizing abilities may be shaped in early childhood by the
low salience of social stimuli (Carruthers, 1996; Sasson, Turner-Brown,
Holtzclaw, Lam, & Bodfish, 2008). Sasson et al. (2008) found that
children with ASD showed decreased gaze time at social objects such
as faces, and increased gaze time at objects that are frequently the
subject of ASD special interests (“high-autism-interest” stimuli), such
as trains, vehicles, and electronics.
The concept of a continuum of special interests is consistent with
social scientists' conception that the domains of human knowledge
range along dimensions of compact/diffuse and rule-governed vs. open-
ended (Simonton, 2009; Toulmin, 1972). We will summarize these
dimensions with the term granularity. Fields of knowledge (interests)
with small granularity are predictable and rule-based, such as the
sciences. The opposing end of the continuum is comprised of large
granularity interests, which, like diffuse domains and “soft” disciplines,
involve few rules and are less definable. Fields of knowledge with large
granularity require the ability to understand broadly defined domains,
which involve few rules. The social environment is the earliest large
granularity domain to which a child is exposed, and mentalizing abilities
develop so that the individual can navigate the social world. High
mentalizing ability may thus represent a cognitive style that is most
attracted to large granularity interests.
The granularity concept provides a means of explaining why an
individual develops interests in particular domains, as the appeal
of phenomena at different granularities would appeal to different
cognitive styles. We propose that each individual has a unique cognitive
“comfort zone,” or preferred granularity, which describes the range of
granularities most attractive to a person with specific cognitive abilities,
such as systemizing and mentalizing. Existing evidence for a continuum
includes Nettle's finding (2007) that NT individuals' scores on the
Systemizing Quotient correlated positively with technological interests,
which are lawful and of small granularity. Billington, Baron-Cohen, and
Wheelwright (2007) found that strong systemizing ability and weak
104 C.L. Caldwell-Harris, C.J. Jordan / Learning and Individual Differences 29 (2014) 98–105

Fig. 3. A proposed continuum of special interests exists ranging the autism-neurotypical continuum. The special interest continuum is defined by systemizing and empathizing or
mentalizing ability. Individuals with autism spectrum disorder, which is also characterized as the extreme male brain, have interests that are highly systemizable and require little
mentalizing ability. Autism spectrum interests are also more intense, specific, and numerous than neurotypical interests. Conversely, neurotypical individuals have interests that may
require high mentalizing ability but less systemizing ability. Neurotypical interests are also less intense, broader, and fewer in number than autism spectrum interests. Female interests
lie on the extreme neurotypical end of the continuum, as neurotypical females tend to have higher empathizing ability than males. Males, who are better systemizers than females but
have lower empathizing ability, would have interests falling on the mid- to end-range of the continuum depending on the strength of their systemizing and empathizing skills.

empathizing ability predicted academic study in the physical sciences
(e.g., mathematics, physics, engineering, etc).
In our study, the reported intensity of interests in large granularity
domains, such as food and drink (which may heavily involve social
interaction) correlated with mentalizing ability. High systemizing
correlated not only with intense small granularity interests but also
with having more intense interests overall, as well as a high total number
of interests. Individuals with large granularity interests may have fewer
interests overall than those with small grain interests, because large
granularity interests are broadly defined and may encompass a wider
range of topics in themselves. The extreme large granularity end of the
continuum represents the cognitive comfort zone of individuals with
high mentalizing ability. NT interests, particularly those of females, are
predicted to occur more frequently on the large granularity end,
as females tend to have higher mentalizing abilities and develop
mentalizing earlier than NT males (Frith & Frith, 2003). The interests of
NT individuals with varying propensities for systemizing and mentalizing
may occur in the middle range of the continuum, but would not
otherwise be labeled as restricted or obsessive.
4.3. Individual variation in cognitive styles
What causes individual differences in systemizing and mentalizing
ability? Previous work has implicated several brain regions in the
development and maintenance of mentalizing, including the mirror
neuron system, the temporo-parietal junction, temporal poles, and
medial prefrontal cortex (Frith & Frith, 2003). The functioning of these
areas can vary depending on many genetic and environmental factors.
Fetal testosterone is thought to be one cause of high systemizing
ability. Fetal testosterone is elevated in boys compared to girls, and
high fetal testosterone was found to correlate with systemizing ability
for both genders (Auyeung et al., 2006). High fetal testosterone also
correlated with boys having a number of special interests, while low
fetal testosterone correlated with high quality social relationships in
both boys and girls (Knickmeyer, Baron-Cohen, Raggatt, & Taylor, 2005).
Another cause of high systemizing ability may be a heightened
attention to detail, as details in a system must be tracked should they
contribute to input-operation–output relationships (Baron-Cohen et al.,
2009). Attention to detail, in turn, could arise from hypersensitivity
to various sensory stimuli, which might affect low-level information
processing and result in the development of interests in unusual domains
(Baron-Cohen et al., 2009). Mottron, Dawson, Soulières, Hubert, and
Burack (2006) propose that hypersensitivity may result from under-
connectivity or local-overconnectivity between neurons in the brain.
The under-connectivity hypothesis suggests that the ASD brain lacks
long-range neural connections responsible for high-level cognitive
processes (Mottron et al., 2006). The failure to develop integrated
connections might lead to hyperspecialization in some brain areas in
the form of local overconnectivity, which could contribute to the
abnormal sensory processing that causes hypersensitivity.
4.4. Implications for education
Understanding special interests is crucial for educators, since
students find their interests inherently rewarding, and educators can
bring them into individualized lesson plans. The continuum model
suggests that individuals with ASD, and males, on average, are more
likely to have small granularity interests, associated with high system-
izing ability. This understanding could be used in the classroom or
work environment, where large granularity topics can be gradually
introduced and related to preexisting interests. For example, children
or adults may be interested in machines and technology, but could be
required to study history, a more “diffuse” domain, for a school or a
workplace requirement. Initial learning could first be stimulated in the
context of small granularity subtopics, such as a discussion of how
technology propelled historical events. Large granularity abstract
concepts could then be progressively added to the curriculum. The
continuum model also suggests that females, on average, tend to have
large granularity interests, associated with high mentalizing ability.
This understanding could be used to adapt educational materials for
teaching more compact small granularity topics, such as mathematics.
Work by Gredlein and Bjoyklund (2005) suggests that even minor
educational tailoring can significantly improve individual performance:
while boys were initially more likely to use tools to retrieve a toy, girls
performed at the same level once they were given verbal hints by
the experimenter, and generalized their learning to a second tool-use
session. It is interesting to note that verbal hints may have capitalized
on girls' mentalizing ability.
The strength-based continuum model of interests can inform career
placement, degree selection, or aptitude testing (Armstrong, 2011;
 C.L. Caldwell-Harris, C.J. Jordan / Learning and Individual Differences 29 (2014) 98–105
Gardner, 2006; Grigorenko & Stemberg, 1997). Understanding the
nature and career potential of special interests may provide support
and relief for parents who are frustrated with their child's consuming
interests. More generally, the continuum approach lends further insight
into individual cognitive differences, and may help us better understand
why these abilities develop. Finally, the special interests continuum
provides a more complete understanding of the deep interests that
motivate people, bring richness to our lives, and often lead to new
scientific and technological advances.
References
Armstrong, T. (2011). The power of neurodiversity: Unleashing the advantages of your
differently wired brain. Cambridge, MA: PerseusBooks.
Asperger, H. (1991). “Autistic” psychopathy in childhood. In U. Frith (Ed., Translator),
Autism and Asperger Syndrome (pp. 37–92). Cambridge: Cambridge University Press.
Attwood, T. (1998). Asperger's syndrome: A guide for parents and professionals. London:
Jessica Kingsley Publishers.
Attwood, T. (2003). Understanding and managing circumscribed interests. In M. Prior
(Ed.), Learning and behaviour problems in Asperger syndrome (pp. 126–147). New
York: Guilford Press.
Auyeung, B., Baron-Cohen, S., Chapman, E., Knickmeyer, R., Taylor, K., & Hackett, G.
(2006). Foetal testosterone and the child systemizing quotient (SQ-C). European
Journal of Endocrinology, 155, 123–130.
Baron-Cohen, S., Ashwin, E., Ashwin, C., Tavassoli, T., & Chakrabarti, B. (2009). Talent in
autism: Hyper-systemizing, hyper-attention to detail and sensory hypersensitivity.
Philosophical Transactions of the Royal Society of London. Series B: Biological Sciences,
364, 1377–1383.
Baron-Cohen, S., Richler, J., Bisarya, D., Gurunathan, N., & Wheelwright, S. (2003). The
systemizing quotient: An investigation of adults with Asperger's syndrome or high
functioning autism, and normal sex differences. Philosophical Transactions of The
Royal Society, 358(1430), 361–374.
Baron-Cohen, S., & Wheelwright, S. (1999). ‘Obsessions’ in children with autism or
asperger syndrome: Content analysis in terms of core domains of cognition. British
Journal of Psychiatry, 175, 484–490.
Baron-Cohen, S., Wheelwright, S., Hill, J., Raste, Y., & Plumb, I. (2001a). The ‘reading the
mind in the eyes’ test revised version: A study with normal adults, and adults with
Asperger syndrome or high-functioning autism. Journal of Child Psychology and
Psychiatry, 42(2), 241–251.
Baron-Cohen, S., Wheelwright, S., Skinner, R., Martin, J., & Clubley, E. (2001b). The autism
spectrum quotient (AQ): Evidence from Asperger syndrome/high functioning autism,
males and females, scientists and mathematicians. Journal of Autism and Developmental
Disorders, 31, 5–17.
Bashe, P. R., & Kirby, B.L. (2001). What Asperger syndrome looks like. In B.L. Kirby, & P. R.
Bashe (Eds.), The OASIS guide to Asperger syndrome: Advice, support, insight, and
inspiration. New York: Crown Publishers.
Billington, J., Baron-Cohen, S., & Wheelwright, S. (2007). Cognitive style predicts entry
into physical sciences and humanities: Questionnaire and performance tests of
empathy and systemizing. Learning and Individual Differences, 17, 260–268.
Carruthers, P. (1996). Autism as mind-blindness: An elaboration and partial defense. In P.
Carruthers, & P. K. Smith (Eds.), Theories of Theories of Mind (pp. 257–273).
Cambridge: Cambridge University Press.
Constantino, J. N., & Todd, R. D. (2003). Autistic traits in the general population: A twin
study. Archives of General Psychiatry, 60, 524–530.
DeLoache, J. S., Simcock, G., & Macari, S. (2007). Planes, trains, automobiles—and tea sets:
Extremely intense interests in very young children. Developmental Psychology, 43,
1579–1586.
Fecteau, S., Mottron, L., Berthiaume, C., & Burack, J. A. (2003). Developmental changes of
autistic symptoms. Autism, 7(3), 255–268.
Frith, U., & Frith, C. D. (2003). Development and neurophysiology of mentalizing.
Philosophical Transactions of the Royal Society of London. Series B: Biological Sciences,
358, 459–473.
105
Frith, U., Morton, J., & Leslie, A.M. (1991). The cognitive basis of a biological disorder:
Autism. Trends in Neurosciences, 14(10), 433–438.
Gardner, H. (2006). Five minds for the future. Boston, MA: Harvard Business School Press.
Grandin, T., & Duffy, K. (2008). Developing talents: Careers for individuals with Asperger
syndrome and high-functioning autism. Shawnee Mission, KS: Autism Asperger
Publishing Company.
Gredlein, J. M., & Bjoyklund, D. F. (2005). Sex differences in young children's use of tools in
a problem-solving task. Human Nature, 16(2), 211–232.
Grigorenko, E. L., & Stemberg, R. J. (1997). Styles of thinking, abilities, and academic
performance. Exceptional Children, 63, 295–312.
Hutt, S. J., Hutt, C. C., Lee, D.D., & Ounsted, C. C. (1965). A behavioural and electro-
encephalographic study of autistic children. Journal of Psychiatric Research, 3(3),
181–197.
Jackson, L. (2002). Freaks, geeks and Asperger syndrome: A user guide to adolescence.
London England: Jessica Kingsley Publishers.
Jordan, C. J., & Caldwell-Harris, C. L. (2012). Understanding differences in neurotypical and
autism spectrum special interests through internet forums. Intellectual and
Developmental Disabilities, 50(5), 391–402.
Klin, A., Danovitch, J. H., Merz, A.B., & Volkmar, F. R. (2007). Circumscribed interests
in higher functioning individuals with autism spectrum disorders: An exploratory
study. Research & Practice for Persons with Severe Disabilities, 32(2), 89–100.
Knickmeyer, R., Baron-Cohen, S., Raggatt, P., & Taylor, K. (2005). Foetal testosterone, social
cognition, and restricted interests in children. Journal of Child Psychology and
Psychiatry, 45, 1–13.
Mazzocco, M., Pulsifer, M., Fiumara, A., Cocuzza, M., Nigro, E., Incorpora, G., et al. (1998).
Brief report: Autistic behaviors among children with fragile x or Rett syndrome:
Implications for the classification of pervasive developmental disorder. Journal of
Autism and Developmental Disorders, 28, 321–328.
Mercier, C., Mottron, L., & Belleville, S. (2000). A psychosocial study on restricted interests
in high-functioning persons with pervasive developmental disorders. Autism, 4(4),
406–425.
Moore, V., & Goodson, S. (2003). How well does early diagnosis of autism stand the test of
time?: Follow-up study of children assessed for autism at age 2 and development of
an early diagnostic service. Autism, 7(1), 47–63.
Mottron, L., Dawson, M., Soulières, I., Hubert, B., & Burack, J. (2006). Enhanced perceptual
functioning in autism: An update, and eight principles of autistic perception. Journal
of Autism and Developmental Disorders, 36(1), 27–43.
Mueller, F., Agamanolis, S., & Picard, R. (2003). Exertion interfaces: Sports over a distance
for social bonding and fun. Proceedings of CHI (pp. 561–568). Fort Lauderdale, FL:
ACM Press.
Nettle, D. (2007). Empathizing and systemizing: What are they, and what do they
contribute to our understanding of psychological sex differences? British Journal of
Psychology, 98, 237–255.
Plank, A., & Yellow Sneaker Media, L. L. C. (2004). WrongPlanet.net: The online resource and
community for autism and Asperger's. http://www.wrongplanet.net/forums.html
(Retrieved March 17, 2010).
Sasson, N. J., Turner-Brown, L., Holtzclaw, T. N., Lam, K. S. L., & Bodfish, J. W. (2008).
Children with autism demonstrate circumscribed attention during passive viewing
of complex social and nonsocial picture arrays. Autism Research, 1, 31–42.
Simonton, D. K. K. (2009). Varieties of (scientific) creativity: A hierarchical model of
domain-specific disposition, development, and achievement. Perspectives on
Psychological Science, 4, 441–452.
Sturm, H., Fernell, E., & Gillberg, C. (2004). Autism spectrum disorders in children with
normal intellectual levels: Associated impairments and subgroups. Developmental
Medicine & Child Neurology, 46, 444–447.
Toulmin, S. (1972). Human understanding, volume 1: The collective use and development of
concepts. Oxford: Clarendon Press.
Turner, M. (1997). Towards an executive dysfunction account of repetitive behavior in
autism. In J. Russel (Ed.), Autism as an Executive Disorder (pp. 57–100). Oxford: Oxford
University Press.
Wheelwright, S., Baron-Cohen, S., Goldenfeld, N., Delaney, J., Fine, D., Smith, R., et al.
(2006). Predicting autism spectrum quotient (AQ) from the systemizing quotient-
revised (SQ-R) and quotient (EQ). Brain Research, 1079, 47–56.
Winter-Messiers, M. (2007). From tarantulas to toilet brushes: Understanding the special
interest areas of children and youth with Asperger's syndrome. Remedial and Special
Education, 28, 140–152.